Enterprise Data Governance
Aligning Data Quality Management
With Your Data Governance Program
Presented by:
Mark Allen
Sr. Consultant, Enterprise Data Governance
WellPoint, Inc.
([email protected])
Introduction:
© WellPoint, 2013 June 2013, DGIQ Conference 2
Mark Allen is a senior consultant and enterprise data governance lead at WellPoint, Inc. Prior to WellPoint, Mark was a senior program manager in customer operation groups at both Sun
Microsystems and Oracle Corporation. At Sun Microsystems, Mark served as the lead data steward for the customer data domain throughout the planning and implementation of Sun’s
customer data hub. Mark has more than 20 years of data management and project management experience including extensive planning and deployment experience with customer master
initiatives, data governance, data integration projects, and leading data quality management practices. Mark has served on various customer advisory boards focused on sharing and enhancing MDM and data governance practices. Mark is also co-author of the book:
Master Data Management in Practice: Achieving True Customer MDM (John Wiley & Sons, 2011).
Contact [email protected] visit http://www.mdm-in-practice.com
WellPoint, Inc:
WellPoint is one of the largest health benefits companies in the United States:
• Revenue: $61.7 billion (2012),
• Net Income: $2.6 billion (2012)
• Employees: 43,650
• Nearly 36 million members in its affiliated health plans
• Nearly 68 million individuals served through its subsidiaries.
Presentation Topics
• Aligning Data Governance with IT
Governance and Project Governance
• Building Data Quality Management within
the Data Governance Program
• Creating and Maintaining an Enterprise
Footprint
• Governing the Data Analysis and Quality
Improvement Process
Governance
Data Quality Management
• Aligning Data Governance with IT
Governance and Project Governance
• Building Data Quality Management within
the Data Governance Program
• Creating and Maintaining an Enterprise
Footprint
• Governing the Data Analysis and Quality
Improvement Process
© WellPoint, 2013 June 2013, DGIQ Conference 4
Data Governance
Data Quality Management
Stuck in Neutral
Companies can struggle with achieving their data management
and quality management objectives because of fundamental
organization and process alignment issues, such as:
Not having data governance aligned well with IT governance
and project governance programs
Organizational changes that can slow momentum or
fragment governance and quality management focus
Not working collectively to correct root causes of data issues
and randomly applying point fixes
Different groups having different systems of record or using
different reporting and analytic solutions
Governance
Data Quality Management
From the TDWI Best Practice Report ’Next Generation Master Data Management’ (2012), among the top responses from users surveyed regarding their challenges to MDM success were:
- Lack of cross-functional cooperation
- Coordination with other disciplines (DI, DQ)
© WellPoint, 2013 June 2013, DGIQ Conference
We have hired a consulting firm to do a master data management
assessment. As project manager I need to quickly pull together a current view of our data architecture and data flows but I am finding it hard to get our data architects to commit enough time for this. I have expensive consultants ready and waiting.
I spent weeks creating a quality dashboard for our product marketing group.
When I previewed this with our Sales and Finance teams they disagreed with a number of my calculations and results. They pointed me to other data and metrics but I can’t get clear answers about that data and those calculations.
You would think that there should be just one common version of account codes or country codes, but instead each of our source systems seems to have its own versions causing me a lot of extra time each month to normalize and recheck this data for executive reports I deliver.
I work in our Customer Service. We are continually challenged because our customer data is inconsistent and not centralized. This clearly
impacts the quality of service we deliver to our customers but we can’t seem to get connected with our other groups who can help address this.
Alignment Issues and Frustrations
(images and captions for illustrative purposes only) 6
Data Governance
Data Quality Management
Lack of Alignment
• Relationships and handshakes between the various governance functions are not well established.
• A Data Governance program exists but lacks sufficient authority and reach.
• IT Governance is not transparent and occurs through various boards and architecture review committees.
• Project Governance functions are distributed across business units resulting in planning inefficiencies and overlaps.
• There is no common data quality management strategy and
framework.
• Business terms, metrics, data models, and metadata are inconsistent.
Data Quality Management
?
IT
Governance
Project Governance Data
Governance
Governance
Data Quality Management
Companies are trying to address the alignment
issues with more integrated data governance
and quality management strategies
From the Information Difference Research Study ’How Data Governance links Master Data Management and Data Quality’ (Aug 2010) involving 257 world-wide companies:
• 58% indicated that their plans for implementing data governance will be part of a broad data management initiative involving data quality.
• “Better quality and faster decisions making” was the top response when asked what are the main benefits they expect Data Governance to deliver.
From the Kalido White Paper ’The Role of Data Quality Monitoring in Data Governance’ (Feb 2011) prepared by Jim Harris, Obsessive-Compulsive Data Quality:
“Data governance provides the framework for a proactive approach to data quality, which requires going beyond reactive data cleansing projects, and establishing a pervasive program for ensuring that data is of sufficient quality to meet the current and evolving business needs of the organization.”
© WellPoint, 2013 June 2013, DGIQ Conference 8
Data Governance
Data Quality Management
From the TDWI Best Practice Report ’Next Generation Master Data Management’ (2012), the top reasons for implementing MDM were:
1. Complete views of business entities 2. Sharing data across the enterprise 3. Data-based decisions and analyses 4. Customer Intelligence
5. Operational excellence
• Use Data Governance as the aligning function with the
IT and Project areas.
• Work with IT management and the project planning
teams to define clear charters, roles, and engagement
opportunities.
• Leverage cross-functional collaboration where it already
exists in projects and programs.
• Bring attention to where good governance and quality
management practices are occurring. Build from best
practices.
• Be persistent, be opportunistic, but have patience.
Key to Building Alignment and Collaboration
Governance
Data Quality Management
© WellPoint, 2013 June 2013, DGIQ Conference 10
An Aligned State
[email protected] June 2012, DGIQ Conference 10
• Governance functions have formal relationships and clear charters.
• Actions and decisions are coordinated through
common strategies, processes, and forums.
• Data stewardship is a core competency with support from IT and Project
resources.
• Data Quality Management is driven through Data
Governance but supported by IT and Project
governance functions.
• Data Quality requirements are defined and applied to IT and Business projects.
IT
Governance Project Governance Data
Governance
Charter:
• Coordination of Enterprise Data Governance Practices
• Data Quality Management Polices and Standards
• Defines Data Steward Roles &
Responsibilities
• Controls Enterprise Business Terms and Rules
• Engaged in IT and Project Governance Decisions
Charter:
• IT Strategies, Investments, Tools, Infrastructure
• Information Architecture, System Architecture
• Technical Review and Solutions.
• Technical Support
• Metadata Management Support
• Engaged in Data Governance and Project Governance Decisions
Charter:
• Project Review, Plan, Budgets
• Project Management, Strategies, Roadmap
• Resource Allocation
• Project Testing, Delivery, and Support
• Support of Data Quality Requirements
• Engaged in IT and Data Governance Decisions Data Quality
Management
Data Governance
Data Quality Management
• Aligning Data Governance with IT
Governance and Project Governance
• Building Data Quality Management within
the Data Governance Program
• Creating and Maintaining an Enterprise
Footprint
• Governing the Data Analysis and Quality
Improvement Process
Governance
Data Quality Management
Data Quality Management Needs To Be…
• A key discipline and function of data governance
• Expressed with clear, measureable milestones in a
data governance maturity model
• Supported through data governance and quality
requirements in the solution design process
• Supported by data stewards and IT associate who
are members of the governance team
• Supported by consistent business processes and IT
solutions
© WellPoint, 2013 June 2013, DGIQ Conference 12
Building Data Quality Management within
the Data Governance Program
Data Governance
Data Quality Management
DQM As A Data Governance Function
Our Data Governance program has been established to define an enterprise-wide data governance foundation and on-going program strategy.
Governance Adoption
& Maturity Metrics
Governance Intake and Decision Metrics
Data Quality Metrics
Metadata Management Metrics
Guiding Principles Data Governance
Model
Data Governance
Measurement
Our data is a strategic enterprise asset with people accountable for its management, quality and integrity.
Our Data Governance policies, standards, and quality requirements will be key factors in our Program Management Office (PMO) and
Solution Design (SDLC) processes.
Data Governance Charter
Processes Services Controls Metrics Data
Stewardship
Metadata Management
Quality Management
Functional Model
Governance
Data Quality Management
DQM In A Governance Maturity Model
© WellPoint, 2013 June 2013, DGIQ Conference 14
Level 1 Marginal & Reactive:
Data governance is at best a marginal and non-formalized
practice. There is need for a more formal structure.
Level 2 Defined & Initiated Data governance has been defined and implemented
within an EDG domain structure, process, and
context.
Level 3
Sustainable & Proactive Data governance is an ongoing practice with active processes engaged in the data
and project life cycles for the data domain.
Level 4:
Optimized & Integrated Data governance is a core competency throughout the
enterprise providing a key role in the EIM and MDM
strategies.
Data Steward Leads and sub-teams are in place
Consistent use and control of business terms and reference data
Data models and
dictionaries are maintained and updated
Data life cycle flows exists
Data source to target mapping is well defined
Key data entities and elements have been identified
Data quality is measured and actively reviewed
Data quality improvement initiatives are ongoing
Data governance and quality management policies are cataloged
Governance and quality requirements are part of SDLC process
A formal data governance charter exists
Data ownership and steward roles are defined
DG processes and collaboration sites have been implemented
DG communication and Training exists
Data model s exists for each data domain
Business Glossary and Metadata Repository solutions exists
Governance engagement exists in key projects
Governance activity metrics and maturity measurements exists
Data quality management orientation plans and processes are underway
Data quality analysis has been initiated
Data ownership is well established and data stewardship is a core competency
Data governance is well integrated with enterprise data architecture and information management strategies, plans, and investments
Master Data Management practices exists with data governance and quality management as key disciplines
Data quality improvement roadmaps are well described within an organization’s continual improvement plans
Risk is well managed with HR, Legal, Compliance, and Privacy Office actively engaged in the DG process
Data governance type decisions are occurring in an ad hoc manner in different decision areas
Data ownership is unclear and quality improvement efforts are disorganized.
Business terms lack standardizations and control
Code sets do not have consistent ownership and management
Some data models exist but are not complete or maintained
Enterprise data management strategies or projects are indicating need for data governance
Audit or compliance issues suggest need for formal data governance
Audits raise data issues and mitigation plans
Customer DG
Product DG
Finance DG
Sales DG
Marketing DG
Data Governance
Data Quality Management
DQM Supported In A Solution Design Process
Discovery &
Requirements
Design &
Develop
Test &
Verify Implement Control &
Maintain Initiative
Planning
Solution Design Process
Engaged in data modeling,
data integration
plans, validation rules, and data policy decisions.
Involved in test and verification
efforts.
Governance team sign-off of data quality
and integrity.
Assist with readiness plans and facilitates resolution of
data related issues.
Involved in data quality
control, metrics, monitoring, and change management.
Data Governance
Identifies needs for data
governance involvement.
Creates data governance engagement
plans as needed.
Participation in discovery
and requirement
sessions.
Identifies any data impacts.
Responds to needs for data
analysis, standards,
general guidance, and
quality metrics.
Governance
Data Quality Management
© WellPoint, 2013 16
DQM Supported By Data Stewards & Analysts
Tactical
Focus
Operational
Focus
Metadata Management
Sub-Team
Data Quality Management
Sub-Team
Strategic
Focus
Compliance Management
Sub-Team
Data Governance Domain Team Structure
Data Domain Trustee
Data Governor Data Governor Data Governor
Data Steward
Data Steward
Data Steward
Data Steward
Data Steward
Data Steward
Data Steward
Data Steward
Data Steward Data
Architect
& Data Analyst Data
Governance
Data Quality Management
Requirements from driver(s)
Analyze and review requirements
Evaluate compliance of requirements
Violation?
Submit request to DG for amendment
Amendment possible?
Requirements cannot be fulfilled
Request data analysis and profiling
Drive the design of solution(s)
Evaluate compliance of
solution(s)
Rules, policies &
procedures
Violation?
Rules, policies &
procedures
Submit request to DG for amendment
Amendment possible?
Requirements cannot be fulfilled Y
N
N Y
Y
N Y
N
DQM Supported By Processes and Solutions
Data Quality Dimension Completeness
Status Data Quality Index 12 Month Trend
Validity Consistency Duplication Accuracy
Detailed Reports
Data Definition Quality
Metadata Management
Sub-Team
Data Quality Management
Sub-Team
Compliance Management
Sub-Team Data Domain
Trustee
Data Governor Data Governor Data Governor
Data Steward
Data Steward
Data Steward
Data Steward
Data Steward
Data Steward
Data Steward
Data Steward
Data Steward Data
Architect
& Data Analyst
Metadata Management
Sub-Team
Data Quality Management
Sub-Team
Compliance Management
Sub-Team Data Domain
Trustee
Data Governor Data Governor Data Governor
Data Steward
Data Steward
Data Steward
Data Steward
Data Steward
Data Steward
Data Steward
Data Steward
Data Steward Data
Architect
& Data Analyst
Metadata Management
Sub-Team
Data Quality Management
Sub-Team
Compliance Management
Sub-Team Data Domain
Trustee
Data Governor Data Governor Data Governor
Data Steward Data
Steward Data Steward Data
Steward Data Steward Data
Steward Data Steward Data
Steward Data Steward Data
Architect
& Data Analyst
Governance
Data Quality Management
• Aligning Data Governance with IT
Governance and Project Governance
• Building Data Quality Management within the
Data Governance Program
• Creating and Maintaining an Enterprise
Footprint
• Governing the Data Analysis and Quality
Improvement Process
© WellPoint, 2013 June 2013, DGIQ Conference 18
Data Governance
Data Quality Management
• Ensure there is a common enterprise-wide process for
data governance intake, item tracking, and decision
response.
• Ensure there is an Executive Council or Board of Trustees
forum for addressing enterprise data governance and
quality management strategies as well as other cross-
domain items or issues.
• Create a user friendly collaboration site and use common
enterprise platforms to support the processes and artifacts
of data governance and quality management
• Define a communication model and RACI matrix for
driving consistent communication
Creating and Maintaining an Enterprise
Footprint
Data Governance
Data Quality Management
© WellPoint, 2013 20
Governance Process Flow Example
Cross- Domain
issue?
Yes No
Qualified Request?
Engage Executive Council?
Executive Council Engagement Cross-
Domain Governance Engagement
Decision
Yes Yes No No
Data Governance
Intake Process &
Item Tracking Log
•Issues
•Data Quality
•Policies
•Standards
•Processes
•Support
•Compliance
•Metadata
•Monitoring
Domain Governance
Review
Data Governance Charter Processes Services Controls Metrics
Data Stewardship
Metadata Management
Quality Management
Functional Model
Data Governance
Data Quality Management
Cross-Domain Governance & Quality
Management
Cross-Domain Data Governance & Data Quality Council (Board of Trustees)
Customer Domain
Stewards Governors
Trustee
Program, Project, and Data Consumer Areas
DG PMO Leads
Product Domain
Stewards Governors
Trustee
Finance Domain
Stewards Governors
Trustee
Other Domains
Stewards Governors
Trustee
Analytics
Analytics Teams
Members
IT, Finance,
Legal, HR, Privacy
Other Key Members
Data Governance Intake Process, Issue Tracking, Meeting Facilitation
Data Governance
Data Quality Management
© WellPoint, 2013 June 2013, DGIQ Conference 22
Common Sites and Platforms Supporting
Data Governance & Quality Management
Collaboration Site
Governance Process
Enterprise Platforms
Enterprise Users
Cross- Domain issue?
Yes No
Qualified Request?
Engage Executive Council?
Executive Council Engagement Cross-
Domain Governance Engagement
Decision
Yes Yes
No No
Data Governance
Intake Process &
Item Tracking Log
•Issues
•Data Quality
•Policies
•Standards
•Processes
•Support
•Compliance
•Metadata
•Monitoring
Domain Governance
Review Data Governance Charter Processes Services Controls Metrics
Data Stewardship
Metadata Management
Quality Management Functional Model
Data Governance
Data Quality Management
Target Audiences (Who to communicate to)
How to Communicate (Communication
Channels)
What to Communicate
When to Communicate
Why to Communicate
• Executives
• Sponsors
• Trustees
• Stakeholders
• Data Governors
• Data Stewards
• Business Users
• Subject Matter Experts
• IT Governance
• IT Teams
• Project Governance
• Project Teams
• New Employees
• Newsletters
• SharePoint
• Wiki
• Bulletin boards
• Management communications
• Project meetings
• Departmental meetings
• Tailored messages for targeted audiences
• FAQ
• DG Charter and Objectives
• DG Structure and Teams
• Accomplishments
• Program News and Announcements
• DG Intake Process
• Other relayed processes
• Metrics and Dashboards
• Policies, Standards
• Meeting and decision info
• Training material
• FAQ
• Relevant Industry artifacts about DG
• Daily
• Weekly
• Bi-Weekly
• Monthly
• Quarterly
• Semi annually Annually
• As needed
• Gain trust
• Keep them informed
• Maintain a presence
• To invoke feedback
• Receive suggestions
• Value of DG and Data Management
Define a Communication Model and RACI
Matrix for Driving Consistent Communication
Data Governance Communication Model
RACI Matrix
Governance
Data Quality Management
• Aligning Data Governance with IT
Governance and Project Governance
• Building Data Quality Management within
the Data Governance Program
• Creating and Maintaining an Enterprise
Footprint
• Governing the Data Analysis and Quality
Improvement Process
© WellPoint, 2013 June 2013, DGIQ Conference 24
Data Governance
Data Quality Management
• Focus on data analysis and quality improvement efforts
that have the most benefit to business operations and
analytics.
• Define, maintain, and publish enterprise standard data
validation rules, quality dimension definitions, and
scorecard formats. This will minimize variations in
quality measurement and format.
• Reduce overlapping solutions, resources, vendor tools,
and consulting engagements.
• Time and effort well spent with developing data
governance will translate over time into better data
quality management and less data quality issues.
Governing The Data Analysis and Quality
Improvement Process
Governance
Data Quality Management
© WellPoint, 2013 June 2013, DGIQ Conference 26
Focus on improvements that have the most
benefit to business operations and analytics
8668
Benefit
Cost
High
High Low
Low
Address Validation & Cleansing
Business Term Standardization
Customer Name Standardization
Account Code Cleanup Parts Code Standardization
Duplicate Customer Merge
Country Code Cleanup
Product Description Cleanup
Data Quality Training
Purchase Data Quality Analyzer Tool
Service Code Analysis
Duplicate Contact Cleanup
Parts Taxonomy Consolidation
Data Governance
Data Quality Management
Data validation rules using standard quality
dimension definitions and scorecard formats
Item Item Title Requirement Description Measurement Criteria DQ Dimension
Zip Code Measure overall zip code field
completeness
Completeness
Measure zip code field completeness from each EPDS v2 data source
Completeness
Measure zip codes for valid format Validity Measure zip code field accuracy
within source systems. Accuracy should using the USPS database as the correct zip code reference source.
Accuracy
Each business entity has one and only one tax id.
Uniqueness
Each tax id has a valid format. Validity 3 Facility Type
Code
Need facility type code of each hospital. Each facility should have valid facility type.
Definitions should come from CMS (e.g., acute facility, SNF, VA hospital)
Each facility should have valid facility type. Definitions should come from CMS (e.g., acute facility, SNF, VA hospital)
Validity
4 D&B DUNS Reference
The D&B service provide DUNS numbers that represent unique Business Entity information to validate or augment a company's customer information. Need to use DUNS numbers as a cross-reference to ensure accuracy of customer data.
Percentage matched, percentage of variance, and percentage below match confidence level.
Accuracy Entry and format of zip codes are in source
systems is inconsistent causing zip code completeness and consistency issues in the enterprise data warehouse. More complete, consistent and accurate capture of zip codes data is required. This will have many benefits including creating more accurate location information for many business and customer services.
1
2 Tax ID Each legal business entity in US should have a unique tax id. Need to ensure these are unqiue and valid ids.
DQ Dimension Enterprise Definition
Completeness Completeness is the measure of missing data.
Consistency Consistency is the measure of the expected data values in one data set being equivalent with values in another data set.
Accuracy Accuracy is the measure of how correct the values agree with an identified reference source of information.
Referential Integrity
Referential Integrity is the measure of the condition that exists when all intended references from the data in one column of a table to data in another column of the same or different table is valid.
Uniqueness Uniqueness is the measure of when no entity exists more than once within the same data set.
Duplication Duplication is the measure of duplication existing within or across systems for a particular field, record, or data set.
Timeliness Timeliness is the measure of the degree to which data is available for use in the time frame in which it is expected.
Currency Currency is the measure of the degree to which information is current with the real world that it models. How
"fresh" the data is in relation to possible time related changes. Has been refreshed within a specified period of time.
Validity Validity is the measure of how well the data conforms to attributes associated with the data element such as its data type, precision, format patterns, range, or expected list of values for the field.
Accessibility Accessibility is the measure of being able to access data when it is required.
Credibility Credibility is the measure of the enterprise users trust and confidence in data.
Data Quality Dimension Completeness
Status
Data Quality Index 12 Month Trend
Validity Consistency Duplication Accuracy
Detailed Reports
Data Definition Quality Governance
Data Quality Management
Benefit of aligned strategies and practices
© WellPoint, 2013 June 2013, DGIQ Conference 28
Time
Effort
+
+
_ _
Data Governance
Data Quality Issues
Data Governance
Data Quality Management
We have hired a consulting firm to do a master data management assessment. As project manager I need to quickly pull together a current view of our data architecture and data flows but I am finding it hard to get our data architects to commit enough time for this. I have expensive consultants ready and waiting.
I spent weeks creating a quality dashboard for our product marketing group.
When I previewed this with our Sales and Finance teams they disagreed with a number of my calculations and results. They pointed me to other data and metrics but I can’t get clear answers about that data and those calculations.
You would think that there should be just one common version of account codes or country codes, but instead each of our source systems seems to have their own versions causing me a lot of extra time each month to normalize and recheck this data for executive reports I deliver.
I work in our Customer Service. We are continually challenged because our customer data is inconsistent and not centralized. This clearly impacts the quality of service we deliver to our customers but we can’t seem get connected with our other groups who can help address this.
In Summary, It’s a Journey…
Looking back, I am happy to say
that we have gotten much better
at addressing our data issues
since we have aligned our data
governance and quality
management programs .
Governance
Data Quality Management
Thank You!
Questions?
© WellPoint, 2013 June 2013, DGIQ Conference 30